Internet protocol (IP) is increasingly used to transmit television signals via packet-based standards. A general term used for all packet-based television is called Internet protocol television (IPTV). Packet-based protocols have traditionally exhibited benefits and drawbacks when compared to traditional broadcast television. One of the primary drawbacks to IPTV is the latency when a subscriber switches programs rapidly. IPTV program acquisitions require a number of steps before the video image is displayed to the subscriber. An IPTV program is typically sent to subscribers via a multicast data transmission and thus the subscriber must leave a current multicast group and join a new group to access a different program. In addition, the time associated with processing video frames further increases cumulative latency of switching from one program to another.
Typically, when a subscriber changes programs during the middle of a group of pictures-based (GOP) multicast transmission, the subscriber needs to wait until the beginning of the next GOP and/or I-frame before being able to view the video image. Video images transmitted in a GOP are a continuous cross-referenced set that require certain GOP frames to display any picture. One suggested method to eliminate this limitation is to send a fast burst of data to the subscriber immediately upon receiving the program acquisition request. The fast burst can contain the video frames contained in the GOP prior to the join point thereby allowing the subscriber to view the video when joining the multicast stream mid-GOP. However, the bandwidth requirements for implementing fast burst can stress network resources without a mechanism to increase the efficiency of fast bursts.
Like reference numbers and designations in the various drawings indicate like elements.